Delay Optimal Adaptive Resource Allocation in 6G for UAVs using Quantum Deep Reinforcement Learning

Abstract

Cellular technology is promising for communication between unmanned aerial vehicles (UAVs). However, integrating cellular technology with UAVs is a significant challenge due to limited bandwidth problems in cellular networks when UAVs transmit large data, or operate in real-time applications. One critical way to enhance the performance of cellular networks is to exploit Terahertz (THz) band communications, one of the key enablers for sixth-generation (6G) networks. THz band operates between 0.1 to 10 THz and could offer ultra-wide spectrum resources. Delay is critical in THz communication especially when UAVs are considered as high delay could result in packet loss. This paper proposes an adaptive resource allocation method to optimize the delay of UAVs under highly dynamic environments. Quantum Deep Reinforcement Learning (QDRL) is formulated since it can learn an optimal policy that maximizes the earned reward quickly even in dynamic environments conditions. Simulation results show that the proposed approach obtains good performance and outperforms the conventional method in different UAV power level and signal-to-noise ratio (SNR).